3892
J . Phys. Chem. 1989, 93, 3892-3894
acetone molecular ions retain large amounts of internal energy for more than 30 ws after formation by electron impact. (2) These highly excited ions are the dominant CID reactants a t such low collision energies. (3) These low-energy (0.45-and 0.65-eV) collisions trigger the release (E T) of about 2.9 eV of stored energy into the recoil of the acetone ion and the collider He atom. With reference to the initial ion beam direction, the acetone ions are backward scattered and He atoms are forward scattered in the CM coordinate frame. (4)The recoiling acetone molecular ion rapidly dissociates into acetyl ion and methyl radical with low kinetic energy of separation. We infer from these observations
-
that the collision process triggers the A X conversion of the acetone molecular ion with the release of most (perhaps all) of the stored electronic energy as translational energy. The vibrationally excited ion then dissociates from the ground-state hypersurface. Our continuing research on polyatomic ion CID dynamics will address the question whether electronically excited states are also important intermediates at higher collision energies.
-
Acknowledgment. Financial support of this research by the National Science Foundation, Grant CHE-83 12069, is gratefully acknowledged.
Radical in the Reaction of C2H, with
Spectroscopic Observatlon of the CH,(S'A,) 0 Atoms
J. Peeters,* S. Vanhaelemeersch, J. Van Hoeymissen, R. Borms, and D. Vermeylen Department of Chemistry, K.V. Leuuen, Celestijnenlaan ZOOF, B-3030 Heuerlee. Belgium (Received: February 13, 1989)
-
Direct spectroscopicobservation of the CH2(11Al)radical in C2H2/O/H systems ( T = 295 K) is reported for the first time. Characteristic rotational features of the 'B,(0,14,0) 'A,(O,O,O) excitation spectrum were positively identified by laser-induced fluorescence. The absolute CH2('A1) concentration was estimated by a calibration based on the CH2C0 CH,('A,) + CO photodissociation at 308 nm. The experimental evidence supports formation of singlet CH2 in C2H2/O/H systems by the fast reaction of H atoms with the HCCO radical, known to be the major product of the primary C2Hz reaction.
Introduction The three most recent mea~urementsl-~of the branching fraction of the elementary reaction of C2H2with 0 atoms: CZH2
+0
+H
(la)
CH2(X3B1)+ C O
(lb)
HCCO
-
-+
all yield k l , / k l values in the range 0.60-0.83, at temperatures from 300 to 2500 K. Theoretical predictions also favor reaction la as the dominant channel, independent of t e m p e r a t ~ r e . ~ Thus, it is imperative to incorporate subsequent reactions of the HCCO radical with 0 atoms and H atoms in ethyne oxidation schemes as pointed out by Peeters et al.;' they specifically emphasized the likelihood that the very fast HCCO + H reaction5 leads to C H I in its metastable singlet I I A I state: HCCO
+H
-
HzCCOt
-
CH2(IlA1)
+ CO
(2)
The CH2('AI)radical as opposed to the ground-state X3B1is highly reactive to closed shell molecule;b10 in hydrocarbon flames, where C2H2is an important intermediate, the singlet methylene (1) Peeters, J.; Schaekers, M.; Vinckier, C. J . Phys. Chem. 1986, 90, 6552. (2) Frank, P.; Bashkaran, K.; Just, T. Symp. ([at.) Combust. [Proc.]2Zsr 1987, 885. (3) Peeters, J.; Boullart, W.; Van de Ven, P.; Vinckier, C. XIIIth International Conference on Photochemistry, Budapest 1987; Abstr., p 69. (4) Harding, L. B.; Wagner, A. F. J . Phys. Chem. 1986, 90, 2974. (5) Vinckier, C.; Schaekers, M.; Peeters, J. J. Phys. Chem. 1985.89, 508. (6) Ashfold, M. N. R.; Fullstone, M. A,; Hancock, G.; Ketley, G. W. Chem. Phys. 1981, 55, 245. (7) Langford, A. 0.;Petek, H.; Moore, B. J . Chem. Phys. 1983, 78, 6650.
(8) Canosa-Mas, C. E.; Ellis, M.; Frey, H. M.: Walsh, R. Inf. J . Chem. Kinef. 1984, 16, 1103. (9) Hack, W.; Koch, M.; Wagner, H. Gg.; Wilms, A. Ber. Bunsen-Ges. Phys. Chem. 1988, 92, 674. (10) Hack, W.; Wagner, H. Gg.; Wilms, A . Ber. Bunren-Ges. Phys. Chem. 1988, 92, 620.
0022-3654/89/2093-3892$01.50/0
-
radical will contribute to the formation of higher unsaturated hydrocarbons by fast reactions with C2H2879and with polyacetylenes. We have already reported3," experimental evidence, albeit of an indirect nature, for the formation of singlet CH2 by reaction 2, both in C2H2/O/H and in C 3 0 2 / H systems (with HCCO C O as primary reaction products in the latter). The evidence is based on the observed strong reduction of molecular beam mass spectrometric (MBMS) CH2 signals upon addition of CHI, which at 300 K is quasi-inert toward CH2(3Bl),but highly reactive toward CH2(1A!).6*7 The prime objective of the present investigation was to obtain conclusive proof for the presence of CH2(IA,) in C 2 H 2 / 0 / H systems by laser-induced fluorescence. The second aim was to evaluate its concentration and to compare it with the value predicted by the proposed formation scheme.
+
Experimental Section A schematic drawing of the equipment is shown in Figure 1. Ethyne ( E 0.01 mbar) was reacted with O / H mixtures (both in the 0.0024.01 mbar range), in a 2.8 cm diameter quartz flow reactor at T = 295 K, ptotN 1 mbar (He as diluent gas), and flow speed u E 1000 cm s-I. The atoms were generated by a microwave discharge through 0 2 / H 2 / H e mixtures; dissociation yields were measured by the MBMS technique. Ethyne was added through a movable central injector tube. A baffled probe laser beam (0.6 cm diameter) passes 1 cm downstream of the reactor mouth at right angles. The pulsed probe beam is generated by a tunable dye laser pumped by a Nd:YAG pump laser (Quantel Datachrome 5000 system). The characteristics are E , N 10 mJ/pulse at X = 585-593 nm, T~ = 13 ns, bandwidth = 0.1 cm-I, and vW = 10 Hz. The energy density drives ( 1 I ) Peeters, J.; Boullart, W.; Van de Ven, P.; Vinckier, C. Joint Meeting of the French and Italian Sections of the Combustion Institute, Amalfi, 1987: paper 2.12.
0 1989 American Chemical Society
The Journal of Physical Chemistry. Vol. 93, No. 10, 1989 3893
Letters FLOW REACTOR
d
r
EXCIMER BEAM ENTRANCE WINDOW
I
ii
VI
c
a
E W
5eb. 300
589 075 I
Figure 1. Schematic illustration of the equipment. The excimer laser beam (not shown) is perpendicular to the dye laser beam and to the fluorescence collection optics axis.
-
the C H 2 GIBI %'Al transition well into saturation. Fluorescence, a t right angles with the probe beam, is focused through a RG-610 high-pass filter onto a Hamamatsu R 928 PM tube. The signal is fed to an S R S 280 boxcar integrator; data acquisition/reduction is achieved by a Macintosh I1 through a Nat. Instr. NB-D10-24 interface supported by Labview software. Reference CH2(IAI) excitation spectra were obtained by photolysis of ketene at 308 nm using a XeCl excimer laser (Lambda Physik EMG 101 MSC; E , N 60 mJ/pulse) crossing the LIF cell at right angles to both the probe laser beam and to the fluorescence collection axis. The delay between the photodissociation pulse and the probe pulse is controlled by a home-built digital delay generator.'* The time delay (5 pus) and the mixture composition (0.002-0.005 mbar ketene and 0.66 mbar helium) ensure a CH?('AI) rotational energy distribution close to 295 K thermal equlllbr~um.'~
Results and Discussion The most intense lines observed in the excitation spectrum of CH2('Al) generated by 308-nm ketene photolysis correspond to the 'B1(0,14,0)404 1A1(0,0,0)414transition at 590.71 nm and the (0,14,0)404 (0,0,0)312transition at 589.37 nm, in agreement with absorption measurements by Petek et al.I4 At dye laser wavelengths fixed at either of the above values, weak fluorescence was also observed in C2H2/O/Hsystems, with a signal level about 2 orders of magnitude lower than obtained in the ketene photodissociation experiments. I t was ascertained that the signals were not due to CH2generated by probe-laser-induced multiphoton dissociation of ketene formed in the C2H2/O/H reaction: no fluorescence was observed when ketene at p N 0.05 mbar was irradiated by the probe laser only. Next, all experimental circumstances and instrumental factors affecting the S / N ratio of the C2H2/O/H system signals were optimized carefully, Le., the mixture composition, the reaction time (determined by the distance between the C2H2-O/H mixing point and the probe laser beam), the fluorescence collection optics, the PM voltage (800 V), the boxcar gate delay (50 ns effective), and the gate width (350 ns). After the optimization, the two strongest lines of the CH2('AI) excitation spectrum could be clearly identified when scanning the dye laser wavelength at l C 3 nm s-I, the lowest possible scan rate. Multiple scans (10-20) were then made over two short spectral regions comprising characteristic groups of CH2('A1) excitation lines including the two most prominent peaks mentioned above; the regions chosen were 589.075-589.525 and 590.625-590.775 nm. The multiple scans were summed, after slight wavelength shifts (1-10 X lo-' nm) on the basis of the 404 -. 3 1 2 or 404 414 line maxima locations as observed for the individual scans; such shifts were necessitated by dye laser wavelength drift.
--
-
(12) Pardon, A.; Peeters, J. Reu. Sci. Instrum. 1988, 59, 2499. (13) Peeters, J.; Borms, R.;Vanhaelemeersch, S.; Van Hoeymissen, J.;
Vermeylcn, D., unpublished results. (14) Petek, H.; Nesbitt, D. J.; Darwin, D.C.; Moore, B. J . Chem. Phys. 1987.86, 1172.
589,525 590.625
590. i 7 5
URUELEHGTH (Ml)
Figure 2. Upper trace: LIF excitation spectra for a C2H2/O/H/He mixture 0.008/0.011/0.003/1.25 mbar; reaction time = 2 ms. Lower trace: parts of the 'B, 'Al excitation spectrum of CH2(IAI)generated by 308-nm photodissociation of ketene; (0,14,0) (O,O,O)band; transitions (A) 606 514; (B) 505 +- 41.3; (C)404 312; (D)80s 81s; (E) 303 211; (F)202 212; (G)404 414; (HI303 313; (1) 202 212.
-
-
-
+ -
-
The signal/noise enhancement achieved by the summation procedure brought to light detailed spectral features. As an example Figure 2 shows the sum of 12 scans for a C2H2/0/H/He mixture with initial partial pressures 0.008/0.01 l/O.O03/ 1.25 mbar at an (optimum) reaction time of 2 ms. The origin of the salient rotational structure must undoubtedly be ascribed to CH2('Al), as demonstrated by comparison with the singlet methylene reference spectra shown at the bottom of the figure (rotational transitions are assigned in accordance with Petek et al.I4). An estimate of the singlet CH2 concentration in C2H2/0/H systems was made by calibration of the signal strengths against the values obtained for CH2('Al) formed by laser photolysis of ketene. The signal strength for a 0.004 mbar of ketene/0.66 mbar of He mixture was a factor 70 larger than for the C2H2/O/H system of Figure 2. With a known photodissociation cross section cm2 at 308 nm15 (yielding solely singlet CH2I6)?a of 3 X measured excimer laser pulse energy of 60 mJ, and a beam section of 2 cmz, the fraction of ketene dissociated to CH2('AI) is 1.4 X with 60% observed to remain after 5 ps (decay time constant = 10 ps). Of course, one should take into account the ratio of the relative populations of the (O,O,O) vibration state for singlet CHI generated by 308-nm ketene photolysis on the one hand and singlet CHI formed in C2H,/O/H systems by reaction 2 on the other hand. According to Nesbitt et al.," the powerful "separate statistical ensemble" (SSE)model of Wittig et a1.,I8 applicable to dissociation processes without exit barrier, predicts a (O,O,O)population of 70% for the 308-nm photolysis of ketene; our SSE calculation yields a (O,O,O) population of only 22% for singlet C H 2 produced by reaction 2, the difference being due to the higher exoergicity of the latter process (see below). The lifetime of CH2('A1) in the C,H,/O/H system of Figure 2 is about 5 ps (governed mainly triplet intersystem crossing collision with He bath by singlet gas atoms6*') implying about 50 collisions, sufficient to ensure rotational relaxation, but far too low for vibrational relaxation of (l,O,O), (O,l,O), or (0.0.1) states to the ground state. Disregarding the combined differences in the effective fluorescence volumes and 'B, fluorescence yields between the two systems (estimated to be 140%), one finally obtains an approximate total CH2(IAI) concentration in the C 2 H 2 / H / 0 system of Figure 2 of 3.8 X lo9 molecules ~ m - ~ . On the other hand, the singlet methylene concentration in the Figure 2 mixture was computed assuming CH2('AI) formation
-
(15) Laufer, A. H. J . Am. Chem. SOC.1971, 93, 61. (16) Hayden, C. C.; Neumark, D. M.; Shobatake, K.;Sparks, R. K.; Lee, Y.T.J . Chem. Phys. 1982, 76, 3607. (17) Nesbitt, D. J.; Petek, H.; Foltz, M. F.; Filseth, S. V.;Bamford, D. J.; Moore, B. J . Chem. Phys. 1985, 83, 223. (18) Wittig, C.; Nadler, I.; Reisler, H.; Noble, M.; Catanzavite, J.; Radhakrishnan, G. J . Chem. Phys. 1985, 83, 5581.
J . Phys. Chem. 1989, 93, 3894-3896
3894
by reaction 2 with a rate constant values of k2 = 1.3 X cm3 molecule-’ s-l. The rate constants for the pertaining C2H2-HCCO- and CH2(’AI) reactions were taken from the current lite r a t ~ r e ; ’ -k~values for the less important singlet CHI removal processes by 0 and H atoms were estimated to be 1.3 X and 2.5 X cm3 molecule-I s-’, respectively. The model predicts [CH2(IA,)] N 4.3 X lo9 molecules cm-3 at its t = 2 ms maximum, Le., nearly equal to the (approximate) measurement. An important additional observation is that a t short reaction times ( 1 2 ms) the CH2(’A!) signals vanish upon removing hydrogen from the reaction mixture; this lends support to the proposed singlet C H 2 formation reaction 2 and rules out direct formation by the primary C2H2+ 0 reaction. As anticipated above, the likely mechanism of reaction 2 is an HCCO + H combination process leading to vibrationally excited ketene in the singlet ground state which then rapidly fragments into CH2(’A,) CO. Thus, the mechanism is quite similar to that for ketene photodissociation at 308 where CH2 is formed solely in the singlet state.I6 There is, however, a difference in the vibrational energy content of the CH2COt(’Al) molecule: in the 308-nm photolysis process it is 93 kcal/mol whereas in the
+
chemical activation reaction 2 it is about 105 kcal/mol, using AHf(HCC0) = 42 kcal/m01’~and AHr(CH2CO) = -1 1 kcal/ mol.20 Thus in the latter case the lifetime for fragmentation will be even shorter and hence singlet triplet intersystem crossing even less probable. Also, the excess energy available to the fragments is only 7 kcal/mol in the 308-nm photolysis,16*”whereas it is 19 kcal/mol for reaction 2; as a consequence, in the latter process a smaller fraction of the singlet CH2 will arise in the (O,O,O) vibrational ground state.
-
Acknowledgment. The work is sponsored by the Inter-University Institute for Nuclear Sciences (I.I.K.W.), Belgium, under a joint project with I.M.E.C., Leuven (contract No. 4.0002.85), and by the Nato Scientific Affairs Division (Contract No SA.5-2-05 RG (85/0756) 1423/85/TT-A1). J.P. is a Research Associate of the National Fund for Scientific Research (N.F. W.O.), Belgium. (19) Oakes, J. M.; Jones, M. E.; Bierbaum, G. M.; Ellison, G. B. Int. J . Chem. Kinet. 1983, 15, 83. (20) Nuttall, R. L.; Laufer, A. H.; Kilday, M. V. J . Chem. Thermodyn. 1971, 3, 167.
Persistent Spectral Hole Burning of Molecular Clusters In a Supersonic Jet Robert J. Lipert and Steven D. Colson* Sterling Chemistry Laboratory, Yale University, New Haven, Connecticut 0651 1 (Received: February 15, 1989; In Final Form: April 3, 1989)
A spectroscopic technique is presented that applies the principles of persistent spectral hole burning to the study of weakly bound clusters formed in a supersonic jet expansion. This approach allows one to distinguish unambiguously between homogeneous and heterogeneous structure in the electronic spectra of mass-selected clusters. Moreover, it makes possible the observation of the electronic spectrum of a single isomer which would otherwise be embedded in the spectra of other isomers. The technique is tested using a mixture of cis and trans isomers of m-cresol. It is then used to establish that the spectrum of the phenol-ethanol hydrogen-bonded complex, in the region of the SI state origin, is due to only a single isomer. The reassignment of the rich vibronic structure near the origin is discussed. Potential applications of this technique in other areas of spectroscopy are briefly considered.
Introduction The ultraviolet-visible absorption spectra of even the simplest van der Waals and hydrogen-bonded clusters formed from polyatomic molecules are often quite complex.’ This spectral congestion may be due to the activity of low-frequency vibrations involving the cluster bonds, the presence of clusters of various sizes with electronic states of similar energy, or the existence of a number of isomeric forms of clusters of the same composition. Multiphoton ionization mass spectrometry can sometimes unfold the overlapping spectra of species with different masses when it is possible to avoid fragmentation of the clusters. The result is then a simplification of the spectra and an unambiguous assignment of the stoichiometry of the clusters responsible for the spectra.2 However, these spectra may still contain many peaks and there remains the uncertainty of whether this structure is homogeneous, Le., due to low-frequency intermolecular vibrational modes of a single cluster, or heterogeneous, Le., due to overlapping spectra of various cluster isomers. One means of resolving this question is through rotational analyses. However, only a few structures of clusters containing large molecules have been obtained from rotationally resolved ~ p e c t r a . This ~ is in part because (1) See, for example, Law, K. S.; Bernstein, E. R. J . Chem. Phys. 1985,
82, 2856. (2) Hopkins, J. B.; Powers, D. E.; Smalley, R. E. J . Phys. Chem. 1981, 85, 3739.
0022-3654/89/2093-3894$01.50/0
sub-Doppler methods are required for the generation of these spectra. In response to this problem, we have developed a spectroscopic technique that can resolve the spectrum of a single cluster isomer in the presence of other isomers that absorb in the same region. Likewise, it can unambiguously rule out cluster isomers as a source of spectral congestion. The ability to go beyond the mass-selective spectroscopy of clusters to isomer-selective spectroscopy holds great promise for contributing to the understanding of dynamical processes in clusters such as cluster isomerization! The technique is based on the principles of laser hole burning: in this case applied in the cold, collision-free environment of a supersonic expansion. Our approach exploits the fact that, subsequent to electronic excitation by laser irradiation, relatively few molecules find their way back to the initial rovibronic levels spanned by the bandwidth of the excitation laser. Photophysical processes such as internal conversion, intersystem crossing, emission to other levels, and ionization by absorption of an additional photon as well as photochemical processes such as predissociation and isomerization (3) See, for example, Haynam, C. A.; Brumbaugh, D. V.;Levy, D. H. J . Chem. Phys. 1983, 79, 1581. (4) Even, U.; Ben-Horin, N.; Jortner, J. Phys. Reu. Lett. 1989, 62, 140. (5) See, for example, Persistent Spectral Hole-Burning: Science and Applications; Moerner, W . E., Ed.; Topics in Current Physics 44; SpringerVkklag: Berlin, 1988.
0 1989 American Chemical Society
